Article Text

Download PDFPDF

Why autonomy needs help
  1. Arthur L Caplan
  1. Correspondence to Arthur L Caplan, Sidney D Caplan Professor of Bioethics, University of Pennsylvania, Department of Medical Ethics, 3401 Market St. Suite 320, PA 22412, USA; caplan{at}


Some argue that to be effective in healthcare settings autonomy needs to be strengthened. The author thinks autonomy is fundamentally inadequate in healthcare settings and requires supplementation by experience-based paternalism on the part of doctors and healthcare providers.

  • Bioethics

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Is bioethics too fervent a believer in the power of personal autonomy? Neil Levy provides a considerable number of reasons to think so in reviewing emerging knowledge about the foibles, biases and limits of human reason.1 ,2 He is right to conclude that there is a lot of evidence that ought to undermine bioethical faith in the power of raw self-determination.

If our brains were not so faulty we would, as Levy notes, be making better choices about healthcare. We surely would make choices that truly advance our deeply held, reflective desires rather than to continue smoking, drinking, cancelling appointments, taking our medicine when we feel like it, swallowing who knows what from the health food store, ignoring concerning symptoms, grasping at last ditch nostrums when we are dying and failing to comply with our doctor's recommendations.

Levy argues that once we absorb the recent findings of decision theory, social psychology and the like and realise how deeply flawed our thinking is we ought accept steps, even if they are a bit paternalistic, to help individuals advance their own conceptions of the good. So, counsellors, debriefers and mandatory cooling off periods have, he maintains, a place at the bedside and in the doctor's office even if the subject or patient might not want them.

Levy argues these steps towards the treatment of our congenitally aberrant power of self-determination ought to be accepted by staunch defenders of autonomy. They are motivated by the desire to let us act on our autonomous conceptions of our own good (good paternalism) rather than by the imposition of someone else's notion of what is good for us (bad paternalism).

I am on board Levy's effort to correct our misguided brains and inborn reward systems so that true autonomy may out.3 ,4 But, Levy both understates the case for why autonomy lacks the power to do what bioethicists expect of it and does not provide sufficient remedies to fix all that ails it.

Autonomy is a core value of Western societies but does not hold pride of place in all of them. Thus, the reliance on autonomy to resolve quandaries frequently found in bioethics is not always sensible or defensible. In many cultures, it makes no sense to have ethicists invoking autonomy when the locals do not understand it, do not want it or simply see putting the personal self above all others as arrogant, grimly isolating or misguided.

Even when autonomy is accorded pride of place in resolving contentious choices in healthcare, self-determination is opaque.5 You can consent to a course of action—say hospitalisation or the implantation of a left-ventricular assist device—without fully realising that you are thereby consenting to a huge number of other actions that will of necessity follow. Consent is inherently limited in that no one can actually predict much less comprehend all the risk and benefit details associated with any given choice. However, that is not the end of autonomy's troubles when it comes to healthcare.

A large number of studies have shown that huge percentages of people who give their informed consent to treatment or to their involvement in research do not really understand what they have chosen. Autonomy lives with hope and hope, in the form of the therapeutic misconception, often trumps autonomy.6

Moreover, those involved in giving consent often fail to understand what consent actually means, what ‘research’ means, misremember giving their consent or do not recall the substance of the consent they gave.7 As well, the opportunities given to those in healthcare settings to choose may still be more reflective of professional or corporate rather than personal values.8 Quizzing or challenging consent can, contrary to Levy's hope, further undermine autonomy.9

Autonomy often does not work in healthcare. Our brains are not designed to let us act upon it. In addition, our cultures do not all value it; quizzing and challenging our choices may make us even less autonomous; we bring too much affect and magical thinking along with us as subject or patient; and our basic memory and perceptual skills fail us when the topic is who is going to stick a needle in our arm or give us a brand new pill in our life-and-death fight against cancer.

What we need in healthcare is a return to a bit more medical paternalism. Conceptions of the good, if they are valid, will not wilt when subjected to dissenting opinion. There is nothing wrong with healthcare providers strongly suggesting a course of care, raising their voice so that you hear their message about health promoting activities, or telling you what they would do if it was their mother in that bed.

Strongly suggesting a course of prevention, care or palliation in the face of a patient's expression of a different choice may be paternalistic but it is a projection of what is good based not upon the doctor's values but upon expertise and experience. Both ought to count in resolving choices in healthcare settings. Autonomy, if it is to be taken seriously, ought to be able to stand up to the professional's vigorously expressed fact-based opinion about what is best.

Sadly, too many decisions are presented to patients and subjects as options without recommendations. The doctor or nurse fears venturing any notion of what might be best for fear of trampling on individual autonomy. Whatever bioethics has sold as respect for autonomy is leaving too many competent and able patients and their surrogates unsure, guilty, bereft, floundering or angry. The solution is not to try harder to fix autonomy. Autonomy is too fragile and too often easily undermined by illness to permit the requisite fix. Rather, we ought to try to buffer autonomy with a bit more vocal input from those who are expert and experienced. The best remedy for our ability to delude ourselves about our own self-determination in healthcare is to force those who try to wield it to consider a possible better alternative as seen by those who have been there before and know better the likely consequences of choosing or not choosing a course of action. If the choice remains the same then at least the patient's view of what is best has been truly put to the test.



  • Competing interests None.

  • Provenance and peer review Commissioned; internally peer reviewed.

Linked Articles

Other content recommended for you